Riyadi, Aan Slamet (2026) Pengembangan Model Optimalisasi Manajemen Sumber Daya Kantin untuk Mendukung Program Employee Wellness melalui Prediksi Pola Kunjungan Karyawan Menggunakan Model Time Series SARIMAX (Studi Kasus: PT XYZ). Masters thesis, Institut Teknologi Sepuluh November.
![[thumbnail of 6032241184-Master_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
6032241184-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (4MB) | Request a copy |
Abstract
Investasi pada kesehatan dan kesejahteraan karyawan melalui fasilitas kantin korporat merupakan upaya strategis untuk meningkatkan moral dan produktivitas kerja. Namun, efektivitas layanan ini sering terkendala oleh masalah superestimasi operasional, di mana angka perencanaan (plan) penyediaan layanan sering melampaui realisasi kunjungan aktual sehingga menimbulkan inefisiensi anggaran dan kerugian finansial yang signifikan. Penelitian ini bertujuan untuk membangun model peramalan time series SARIMAX (Seasonal Autoregressive Integrated Moving Average with Exogenous Factors) untuk memberikan proyeksi kunjungan harian secara presisi guna meminimalisir gap antara rencana dan realisasi aktual pada kantin S PT XYZ. Penelitian menggunakan pendekatan kuantitatif berbasis metodologi Box-Jenkins atas 1.035 data kunjungan harian (Juli 2023–April 2026). Tujuh kandidat variabel eksogen diuji melalui grid search dan backward elimination berbasis p value, meliputi periode Ramadan, Idul Fitri, libur sekolah, serta kegiatan internal perusahaan. Akurasi diukur dengan RMSE, MAE, dan MAPE. Model final SARIMAX(1,1,2)(1,1,1)₇ menetapkan tiga variabel eksogen yang signifikan (p<0,001): Ramadan, Idul Fitri, dan libur sekolah; sedangkan kegiatan internal terbukti tidak signifikan. Model mencapai MAPE 10,14% pada kondisi normal (good forecasting) dengan residu white noise. Simulasi menunjukkan reduksi superestimasi 68,8%, setara potensi penghematan sekitar Rp 600,2 juta per tahun. Temuan ini memberikan landasan ilmiah bagi optimasi anggaran kantin sekaligus menjamin keberlanjutan program employee wellness.
===============================================================================================================================
Investing in employee health and well-being through corporate canteen facilities is a strategic effort to enhance workplace morale and productivity. However, the effectiveness of this service is often hindered by operational overestimation, in which the planned (plan) service provision frequently exceeds actual visit realization, thereby creating budget inefficiency and significant financial losses. This study aims to develop a SARIMAX (Seasonal Autoregressive Integrated Moving Average with Exogenous Factors) time series forecasting model to provide precise daily visit projections in order to minimize the gap between plan and actual realization at Canteen S of PT XYZ. The study employs a quantitative approach based on the Box-Jenkins methodology, using 1,035 daily visit records (July 2023–April 2026). Seven candidate exogenous variables were examined through grid search and p-value-based backward elimination, covering the Ramadan period, Eid al-Fitr, school holidays, and internal corporate activities. Accuracy was measured using RMSE, MAE, and MAPE. The final model, SARIMAX(1,1,2)(1,1,1)₇, established three significant exogenous variables (p<0.001): Ramadan, Eid al-Fitr, and school holidays, whereas internal corporate activities proved insignificant. The model achieved a MAPE of 10.14% under normal conditions (good forecasting) with white noise residus. The simulation demonstrated a 68.8% reduction in overestimation, equivalent to potential savings of approximately IDR 600.2 million per year. These findings provide a scientific basis for optimizing canteen budgets while ensuring the sustainability of the employee wellness program.
| Item Type: | Thesis (Masters) |
|---|---|
| Uncontrolled Keywords: | SARIMAX, Superestimasi, Efisiensi Finansial, Kunjungan Kantin, Employee Wellness,Overestimation, Financial Efficiency, Canteen Visits, |
| Subjects: | H Social Sciences > HA Statistics > HA30.3 Time-series analysis |
| Divisions: | 61101-Magister Management Technology |
| Depositing User: | Aan Slamet Riyadi |
| Date Deposited: | 31 Jul 2026 00:44 |
| Last Modified: | 31 Jul 2026 00:44 |
| URI: | http://repository.its.ac.id/id/eprint/140200 |
Actions (login required)
 |
View Item |